Sample Characteristics

The analytic sample contained 14,088 initial visit entries (mean [SD] age = 39.17 [20.84], face-to-face [%] n = 10,474 [74.35%], video [%] n = 3614 [25.65%]), and was predominantly comprised of adults (9412, 66.81%), whites (13,031, 92.50%), non-Hispanics (31,541, 96.12%), females (8557, 60.74%), and those with either Medicare or Medicaid (6848, 48.61%) or Commercial insurance (6830, 48.48%). Additional sample composition can be found in Table 1.

Punctuality

Punctuality was calculated as the appointment time minus check-in time, with negative values reflecting early arrivals and positive values reflecting tardy arrivals.19 Overall sample check-in time demonstrated a mean of -9.67 (SD = 26.12, Range = -463 – 443 min). F2F visits demonstrated a mean check-in time of -10.78 (SD = 26.77, Range = -345 – 443 min), while video visits demonstrated a mean check-in time of -6.44 (SD = 23.87, Range = -463 – 336 min).

Punctuality by Factor

A binary logistic regression was conducted (0 = early/on-time, 1 = late) to evaluate F2F and video service type’s influence on promptness. The model was statistically significant, χ2(1) = 4.33, p = 0.04, suggesting that when compared to F2F, increased video use was associated with a decreased likelihood of late check-in (B = -0.10, SE = 0.05, Exp(B) = 0.91, 95% CI = 0.83 – 1.00). While significant, the model explained < 1% of the variance (Nagelkerke R Square = 0.000465). Primary and exploratory analyses are presented in Table 2.

Table 2 Primary and Exploratory Binary Logistic Regression AnalysesExploratory Analyses

Exploratory binary logistic regressions were conducted for video visits to determine trends in punctuality. Comparison points were selected for each category as the most common presentation of the services as based upon author consensus.

Age

The model was statistically significant, χ2(2) = 18.46, p < 0.001, and suggested that when compared to adult patients, child/adolescent patients had a higher likelihood of late check-in (B = 0.44, SE = 0.10, Exp(B) = 1.55, 95% CI = 1.27 – 1.89). While significant, the model explained less than 1% of the variance (Nagelkerke R Square = 0.01).

Sex

The model was statistically significant, χ2(1) = 4.06, p = 0.04, and suggested that when compared to females, males had a higher likelihood of late check-in (B = 0.17, SE = 0.09, Exp(B) = 1.19, 95% CI = 1.01 – 1.41). While significant, the model explained less than 1% of the variance (Nagelkerke R Square =  < 0.00).

Specialty

The model was statistically significant, χ2(2) = 85.10, p < 0.001, and suggested that when compared to psychiatrist appointments, social worker appointments had an increased likelihood of late check-in (B = 0.58, SE = 0.10, Exp(B) = 1.79, 95% CI = 1.48 – 2.16). Contrastingly, psychologist appointments had a decreased likelihood of late check-in (B = -5.7, SE = 0.11, Exp(B) = 0.57, 95% CI = 0.45 – 0.71). While significant, the model explained less than 4% of the variance (Nagelkerke R Square = 0.04).

Insurance Type

The model was statistically significant, χ2(2) = 14.10, p < 0.001, and suggested that when compared to Medicaid and Medicare, commercial insurance patients had a decreased likelihood of late check-in (B = -0.30, SE = 0.08, Exp(B) = 0.74, 95% CI = 0.63—0.87). While significant, the model explained less than 1% of the variance (Nagelkerke R Square = 0.01).

Diagnostic Classification

The model was statistically significant, χ2(16) = 62.25, p < 0.001; and suggested that when compared to Medical Conditions, Gender Identity and Associated Conditions (B = 1.11, SE = 0.57, Exp(B) = 3.05, 95% CI = 0.99 – 9.38), Disruptive Behavior and Associated Conditions (B = 0.98, SE = 0.37, Exp(B) = 2.66, 95% CI = 1.29 – 5.46), Substance Use and Associated Conditions (B = 0.60, SE = 0.25, Exp(B) = 1.83, 95% CI = 1.12 – 2.98), and Neurocognitive and Associated Conditions (B = 0.83, SE = 0.35, Exp(B) = 2.28, 95% CI = 1.14 – 4.56) each had an increased likelihood of late check-in. Contrastingly, Non-Specific Mental Health Conditions (e.g., “psychological factors affecting morbid obesity,” “martial conflict;” B = -0.59, SE = 0.26, Exp(B) = 0.56, 95% CI = 0.33—0.92), demonstrated a decreased likelihood of late check-in compared to Medical Conditions. While significant, the model explained less than 3% of the variance (Nagelkerke R Square = 0.03).

Ethnicity and Race

Models for ethnicity (p = 0.89) and race (p = 0.82) were nonsignificant.